Home / Ideas / RSI(2) Time-Series Mean Reversion on Index ETFs
RSI(2) Time-Series Mean Reversion on Index ETFs
buy short-term oversold, exit on snap-back, long-only above the trend filter
Larry Connors' 2-period RSI is a heavily-documented short-term reversion rule; SPY backtests (1993–present) show ~9% annual return while invested only ~28% of the time, with >75% win rate on 1–3 day holds. Simple, fully systematic, and robust when gated by a long-term trend filter (only buy dips above the 200-day SMA).
Why it's relevant now
With VIX oscillating in the 15–19 band and periodic risk-off spikes, brief panic sell-offs that snap back within days are frequent — but the trend filter is essential given the mid-June risk-off cluster that hurt unfiltered index reversion.
Universe
Broad index and bond ETFs — SPY, QQQ, IWM, TLT, EFA, EEM (single-asset, no shorting required). Free daily OHLCV.
How it works
When price is above its 200-day SMA and 2-period RSI < 10, go long the ETF; exit when 2-period RSI > 70 or price closes above the 5-day SMA; equal-weight across triggered ETFs.
Expected performance
Research-derived Sharpe estimate: 0.5–0.9. moderate; the ~34% max drawdown in unfiltered SPY tests shrinks materially with the 200-day gate).
Backtest this idea with SignalChain
This is a research lead — not a finished backtest. SignalChain takes an idea like this and runs the whole pipeline inside Claude Code: it researches the concept against academic and practitioner sources, sets benchmarks, writes and lints a VectorBT backtest, runs it, and grades the result PASS/FAIL. One command:
/signalchain Long-only RSI(2) mean reversion on index ETFs (SPY, QQQ, IWM, TLT): buy when 2-period RSI below 10 and price above 200-day SMA, exit on RSI above 70.
Get SignalChain — $49 →
Research & sources
- quantifiedstrategies.com — quantifiedstrategies.com
- stratbase.ai — stratbase.ai
- algotr.substack.com — algotr.substack.com
More mean-reversion and statistical arbitrage ideas
Not financial advice. This page describes a research idea, not a recommendation. Any performance figures are hypothetical, research-derived estimates and are not indicative of future results. SignalChain is a research and educational tool; you are solely responsible for any decisions you make.